Using earth observation data to monitor the Sustainable Development Goals
One sentence summary:
EO data can be used to help quantify 108 indicators from the UN Sustainable Development Goals; the majority of these indicators are environmentally focused.
Summary
Quantifying a nation’s progress with respect to the Sustainable Development Goals (SDGs) can be time-consuming and expensive, particularly when it comes to data collection and analysis. A recent paper in Sustainability by Andreas et al. outlines the role that earth observation data (EO) can play in quantifying the SDGs.
This paper notes that EO data can be used as a proxy dataset for variables that would be too expensive to collect, such as air pollution data or household wealth. They reference two studies that tried to model the global cost of quantifying our progress on the SDGs across every nation. The modelled cost (across two studies) varied between 4.5 billion to 254 billion dollars!
There are, of course, limitations to using EO data, which the author broke down as follows:
Limitations in resolution: This includes spatial, temporal, and spectral. Using EO data in sensitive applications – such as disaster management – can be limiting if the appropriate resolution isn’t available. However, with all the upcoming satellite launches (approximately 200 are planned), this will become less of an issue.
Cost: EO data provided by governmental organizations is free, but the appropriate product may not be available, leading users to commercial sources of EO data. However, this may be cost-prohibitive depending on the study area and time series required for the SDG indicator, demonstrating the need for free and open data.
Inadequate training: The author notes that many countries (particularly those in developing nations) lack the appropriate statistical institutions to using EO data appropriately. The positive news here is the growing number of free, online resources available to fill this gap. One example is the course material developed by NASA, which contains several training sessions focused specifically on SDGs.
How many SDG indicators can use EO data?
The authors of this study used an analysis technique called the “maturity matrix framework” to categorize the level of support that EO data could provide to each SDG indicator. The results were as follows:
Weak EO support for 19 indicators
Partial support for 67 indicators
Strong support for 22 indicators
No support for 139 indicators
EO data was found to provide the most value when trying to quantify SDGs with an environmental focus. However, EO data could still be used as a proxy variable for socioeconomic indicators. For example, satellite remote sensing data of an urban layout (building shape, street orientation) were used to model and predict homicide rates. This study found that homicide rates were more prevalent in areas with a “disordered urban layout.”
Figure 1 provides a breakdown of the maturity matrix analysis. The y-axis provides the number of points given to each indicator. Refer to the paper for a closer look at the methods behind this analysis.
Private companies have recognized the values that EO data can provide, and collaborations have already taken place. Planet Labs developed the "Global Forest Watch" using their very high resolution (VHR) imagery. Maxar Technology also used their VHR data but targeted a socioeconomic SDG. They tried using variables developed from their VHR imagery to understand if they could predict economic well-being.
This paper references numerous research studies that have already used EO data to quantify SDGs through different methods. Although most of these appear to be proof-of-concept exercises, they demonstrate the value that EO data can provide. A few examples are as follows:
Economic well-being: VHR data was used to model economic well-being across Sri Lankan villages. Very High Resolution (VHR) data were used to track car counts, building density, and green space and these variables were used as proxy data to understand the spatial distribution of poverty.
Malaria risk: Population density and proximity to standing water was used to track malaria risk. These maps were used to understand how health care workers should distribute mosquito nets and spray insecticide.
Economic loss: EO optical and radar imagery was used to detect building damage after an earthquake. This can be used to understand economic loss.
Household wealth: Spatial landscape use, building size, and other variables were derived from EO data and used to predict household wealth in western Kenya; model results demonstrated a 62% accuracy.
Andries et al. also note that new indicators may be developed to predominantly use EO data, which should be exciting news for those in the consulting sector.
To conclude, this paper is a comprehensive review of the role EO data can play as nations continue to tackle the SDGs, and I encourage you to read the full publication.
Reference: Andries, A.; Morse, S.; Murphy, R.J.; Lynch, J.; Woolliams, E.R. Using Data from Earth Observation to Support Sustainable Development Indicators: An Analysis of the Literature and Challenges for the Future. Sustainability 2022, 14, 1191. https://doi.org/10.3390/su14031191